Enhanced management zones for precision agriculture
The present invention is a system and method for agricultural management-zone delineation to be done over broad geographic extents without overly-localized field-specific data. The instant innovation guides precision agricultural sampling and management by delineating enhanced management zones based upon remote sensing and artificial intelligence and combining the two with data derived from an existing countrywide soil survey database. In an embodiment, the instant innovation uses artificial intelligence from multiple sources to provide granular zone detail. Output of the present innovation can be aggregated to produce management zone sizes that have a level of uncertainty compatible with the needs of the customer-farmer and implementable given the capabilities of available equipment.
1. A system for optimizing agricultural soil zone attributes comprising:
a data processor in communication with a data server;
a user device capable of displaying data representations to a user;
a first dataset constructed from remote sensing and digital analysis;
a second dataset constructed from collected imagery bands and calculated indices;
a third dataset constructed from a combination of inputs to the first dataset and the second dataset;
wherein the data processor applies artificial intelligence algorithms to the first dataset and outputs a first set of soil zone attributes;
wherein the data processor applies artificial intelligence algorithms to the second dataset and outputs a second set of soil zone attributes;
wherein the data processor applies artificial intelligence algorithms to the third dataset and outputs a third set of soil zone attributes;
wherein the data processor compares the first, second, and third sets of soil zone attributes;
wherein the data processor creates preliminary zone boundaries from the first, second, and third sets of soil zone attributes by (1) maximizing a total number of statistically different soil zone attributes, (2) maximizing inter-zone differences between soil zone attributes, and (3) minimizing within-zone variances;
wherein the data processor calculates a change in at least one set of soil zone attributes from an aggregated zone comprised of the preliminary zone and one or more zone patches and/or inclusions;
wherein the data processor dissolves zone boundaries of the preliminary zone and the one or more zone patches and/or inclusions; and
wherein the data processor delivers an optimized set of soil zone attributes and one or more zone management recommendations to the user.
2. The system of claim 1 , where the remote sensing is achieved with LIDAR.
3. The system of claim 1 , where the first dataset, second dataset, and third dataset are supplemented by SSURGO Soil Parent Material Data.
4. The system of claim 1 where the artificial intelligence algorithms are Deep Learning algorithms, Unsupervised Learning algorithms, or a combination of Deep Learning and Unsupervised Learning algorithms.
5. The system of claim 1 where an evaluation of zone effectiveness incorporates ground truth or Deep Learning algorithm output.
6. A method for optimizing agricultural soil zone attributes with a data processor comprising:
constructing a first dataset using remote sensing and digital analysis;
constructing a second dataset using collected imagery bands and calculated indices;
constructing a third dataset using a combination of inputs to the first dataset and the second dataset;
applying artificial intelligence algorithms with the data processor to the first dataset and outputting a first set of soil zone attributes;
applying artificial intelligence algorithms with the data processor to the second dataset and outputting a second set of soil zone attributes;
applying artificial intelligence algorithms with the data processor to the third dataset and outputting a third set of soil zone attributes;
comparing with the data processor the first, second, and third sets of soil zone attributes;
creating with the data processor preliminary zone boundaries from the first, second, and third sets of soil zone attributes by (1) maximizing a total number of statistically different soil zone attributes, (2) maximizing inter-zone differences between soil zone attributes; and (3) minimizing within-zone variances;
calculating with the data processor a change in at least one set of soil zone attributes from creating an aggregated zone comprised of the preliminary zone and one or more zone patches and/or inclusions;
dissolving with the data processor zone boundaries of zones of the preliminary zone and the one or more zone patches and/or inclusions; and
delivering with the data processor an optimized set of soil zone attributes and one or more zone management recommendations to the user.
7. The method of claim 6 , where the remote sensing is achieved with LIDAR.
8. The method of claim 6 , where the first dataset, second dataset, and third dataset are supplemented by SSURGO Soil Parent Material Data.
9. The method of claim 6 , where the artificial intelligence algorithms are Deep Learning algorithms, Unsupervised Learning algorithms, or a combination of Deep Learning and Unsupervised Learning algorithms.
10. The method of claim 6 , where an evaluation of zone effectiveness incorporates ground truth or Deep Learning algorithm output.